Clinical and analytical evaluation of an enzyme immunoassay for myelin basic protein in cerebrospinal fluid

BACKGROUND: RIA of myelin basic protein (MBP) in cerebrospinal fluid (CSF) is commonly used as a biochemical marker of demyelination in patients with multiple sclerosis (MS). Our aim was to develop a sufficiently sensitive ELISA for MBP and evaluate it clinically in patients with MS. METHODS: The ELISA used anti-bovine MBP antibody coated on plates and biotinylated anti-MBP antibody. The bound antibody complex was quantified with streptavidin-horseradish peroxidase. MBP was determined in CSF from 84 MS patients and 55 patients with other neurological diseases. RESULTS: The respective within- and between-assay CVs were 4.7% and 7.2% at 200 ng/L, and 6. 3% and 8.8% at 2000 ng/L. The detection limit was 30 ng/L. Most of the MS patients with acute exacerbations had markedly increased MBP in the CSF. Longitudinal studies of six MS patients with recurrent exacerbation confirmed this observation. MBP concentrations from 78 MS patients, as tested with our ELISA, correlated well with those obtained by RIA (r = 0.9; P: <0.01), but the detection limit of the ELISA was much lower than that of the RIA. CONCLUSIONS: This convenient ELISA with higher sensitivity than the existing assays is a suitable routine assay that provides a diagnostic indicator of myelin breakdown in the central nervous system; moreover, it is an excellent indicator of MS disease activity.     

Localization of eosinophil granule major basic protein in human basophils

In experiments using an immunofluorescent method to localize human eosinophil granule major basic protein (MBP) in cells and tissues, a small number of cells stained for MBP that subsequently could not be identified as eosinophils. Because the Charcot-Leyden crystal protein, another eosinophil protein, was recently identified in basophils, we tested whether MBP might also be a constituent of blood basophils. Highly purified, eosinophil-free basophil suspensions were prepared using the fluorescence-activated cell sorter (FACS) to sort basophil-containing mononuclear cell preparations stained with fluorescein-conjugated sheep IgG anti-human IgE antibody. Using these FACS-purified basophils, we found that: (a) enrichment for surface IgE-positive cells (greater than 95% basophils) by FACS also enriched for cells staining for MBP by immunofluorescence; (b) MBP appeared to be localized in the histamine-, heparin-containing granules; (c) significant quantities of MBP were measurable by radioimmunoassay (RIA) in freeze-thaw detergent extracts of purified basophils; and (d) RIA dose-response curves for extracts of purified eosinophils and basophils had identical slopes. The MBP content of basophils from normal individuals averaged 140 ng/10(6) cells, whereas purified eosinophils from normal donors and patients with the hyper-eosinophilic syndrome averaged 4,979 and 824 ng/10(6) cells, respectively. MBP was also detected by immunofluorescence and RIA in cells obtained from a patient with basophil leukemia, although the MBP content for basophil leukemia cells was lower than that for normal basophils. We conclude that basophils contain a protein that is immunochemically indistinguishable from eosinophil granule MBP.     

Analytical performance of EMIT cyclosporine assay evaluated.

We evaluated the EMIT Cyclosporine Assay (Syva Co., Palo Alto, CA), using the Cobas-Mira analyzer to assess the precision, accuracy, and analytical recovery from whole-blood samples supplemented with cyclosporine. We also performed comparative analysis of whole-blood samples containing cyclosporine from liver and kidney transplant patients by using EMIT, HPLC, and RIA (IncStar Cyclo-Trac, SP assay). Before assay by EMIT or RIA, cyclosporine was extracted from whole blood with methanol. For the HPLC method, whole blood containing cyclosporine was hemolyzed with 300 mL/L acetonitrile in water; cyclosporine was extracted from the hemolysate with acetonitrile. The within-run and between-run CVs for the EMIT assay of cyclospoprine were 9.9% (means = 72.6, SD = 7.2 micrograms/L; n = 20) and 13.5% (means = 75.0, SD = 10.1 micrograms/L; n = 26) for the low control; 3.5% (means = 194.7, SD = 6.8 micrograms/L; n = 20) and 8.1% (means = 189.0, SD = 15.3 micrograms/L; n = 26) for the medium control; and 7.0% (means = 332.5, SD = 23.3 micrograms/L; n = 20) and 7.1% (means = 340.0, SD = 24.2 micrograms/L; n = 24) for the high control (Bio-Rad, whole-blood controls). Analytical recovery of cyclosporine from drug-supplemented samples averaged 99% for EMIT, 104% for HPLC, and 90% for RIA over a concentration range of 50-500 micrograms/L. Analysis of 196 specimens by HPLC (x) vs EMIT (y) gave the following regression statistics: y = 1.27x + 16.44; IncStar's RIA (x') vs EMIT: y = 1.12x' - 2.50; HPLC vs RIA: x' = 1.10x + 23.87.     

Measurement of myelin basic protein by radioimmunoassay in closed head trauma, multiple sclerosis and other neurological diseases

A double antibody sequential radioimmunoassay for human myelin basic protein (MBP) has been developed. The assay utilizes a rabbit antibody to human MBP and purified rabbit MBP as the radiolabelled antigen. This assay was used to analyze cerebrospinal fluid (CSF) from 22 patients with severe head injury, 61 other cases of various neurological disorders, and 106 normal controls. The results showed that closed head trauma caused moderate to severe elevations in CSF MBP, and elevated CSF MBP was detectable in several diseases which involve CNS myelin.     

Cerebrospinal fluid C-reactive protein in infective meningitis in childhood

The value of cerebrospinal fluid C-reactive protein (CSF CRP) determination as a diagnostic aid in infective meningitis has been investigated in four groups of children. In a "no meningitis" group of 10 children, a median CSF CRP value of 0.08 micrograms/ml was obtained (range 0 to 0.31 micrograms/ml); in a viral meningitis group of 21 children a median value of 0.01 micrograms/ml (range 0 to 3.06 micrograms/ml); in a bacterial meningitis group of 27 children a median value of 9.6 micrograms/ml (range 0 to 31.5 micrograms/ml); and in a tuberculous meningitis group of 18 children a median value of 0.29 micrograms/ml (range 0 to 4.9 micrograms/ml). CSF CRP values in the bacterial meningitis group differed significantly from those of each of the other groups (P less than 0.01), but considerable overlap between the groups detracted from the diagnostic value of the test. In six patients with bacterial meningitis with ambiguous conventional CSF chemistry results, normal CSF CRP values were found. Simultaneous serum CRP was determined in nine patients with tuberculous meningitis and 11 with bacterial meningitis, and the CRP response in both the serum and CSF appears subdued in tuberculous meningitis in comparison with bacterial meningitis. CSF CRP and total protein values were determined intermittently during a 24-hour period in ventricular CSF from two children with tuberculous meningitis who underwent temporary direct ventricular drainage. A considerable and apparently parallel diurnal variation in both values was seen. CSF CRP values have limited application in the etiologic diagnosis of meningitis.     

Differential assay of zidovudine and its glucuronide metabolite in serum and urine with a radioimmunoassay kit

We developed an ancillary procedure for the ZDV-Trac RIA (Incstar) to allow simultaneous determination of both zidovudine (3'-azido-3'-deoxythymidine, ZDV, AZT, Retrovir) and its metabolite, the glucuronide of ZDV (3'-azido-3'-deoxy-5'-O-beta-D-glucopyranuronosylthymidine, ZDVG, GAZT), in human serum and urine. Using the ZDV-Trac RIA, we measured ZDV concentrations before and after ZDVG in samples was hydrolyzed to ZDV by beta-glucuronidase (EC 3.2.1.31); ZDVG concentration was calculated as the difference between the two results. This method enables rapid evaluation of a large number of samples with a total turn-around time of 6 h. The lower detection limit of the RIA was 0.27 micrograms/L; the measurements varied linearly with ZDV concentrations from 0.27 to 217 micrograms/L, with the 50% inhibitory concentration being approximately 10 micrograms/L. Analytical recoveries of inhouse serum and urine controls for both ZDV and ZDVG exceeded 90%. Coefficients of variation (CVs) of serum controls were less than 6% for ZDV and less than 11% for ZDVG; for urine controls, CVs for both ZDV and ZDVG were less than 6%. Results for ZDVG concentrations obtained by HPLC and by the ZDV-Trac RIA system compared well: r = 0.978, slope 1.0, for serum samples, and r = 0.993, slope 1.09, for urine samples.     

High-performance liquid-chromatographic method compared with a modified radioimmunoassay of cotinine in plasma

Cotinine is a sensitive and specific biochemical marker of exposure to cigarette smoke. We describe a simple solid-phase extraction of cotinine from plasma before quantification by HPLC. Extraction recovery was 97.9% +/- 11.0% for plasma concentrations of 5-400 micrograms/L. Baseline separation of cotinine and caffeine was achieved within 11 min of injection onto a C18 reversed-phase column. The mobile phase was citric acid/dibasic potassium phosphate (30 mmol/L each, pH 6.0) containing 100 mL of acetonitrile per liter. Within-day and day-to-day precision (CV) were 4.7% and 8.4%, respectively. We also describe a modification of the Nicotine Metabolite RIA kit (Diagnostic Products Corp.) for quantifying cotinine in plasma. Recovery of cotinine from supplemented plasma was within 10% of the expected value with this RIA kit. Interassay precision averaged 8.1% for samples in the range 50-400 micrograms/L; intra-assay precision averaged 3.6% at 230 micrograms/L and 8.7% at 53 micrograms/L. Correlation between the two methods was RIA = 1.13 HPLC + 14.8 (n = 128, r = 0.957, P less than 0.001). Both methods are technically simple to perform.     

Highly sensitive immuno-assays for the determination of cotinine in serum and saliva. Comparison between RIA and an avidin-biotin ELISA

Two immuno-assay methods (RIA and ELISA) have been developed for the accurate and sensitive measurement of cotinine in human body fluids (serum, saliva). RIA uses [3H]cotinine as antigen and charcoal/dextran for separating cotinine-bound antibodies from the free derivative. Another technique (ELISA) was developed to avoid the use of radio-labelled compounds and to determine cotinine in large populations, including passive or non-smokers who usually present very low concentrations. The two techniques were analytically validated. The detection limit was similar (0.1 micrograms/l) and the precision was better than 10% for both techniques. Non-smoker values ranged from 0.1 to 17 micrograms/l by ELISA and 0.1 to 27.5 micrograms/l by RIA, whereas smoker values ranged from 50 to 1000 micrograms/l (ELISA) and from 70 to 800 micrograms/l (RIA). The comparative analysis of cotinine in 96 human sera revealed a good correlation between the two methods (r = 0.97) and a reliable discrimination between the populations of non-smokers and smokers. As usual, the ELISA is more rapid (4 h 30 min) than the RIA (longer than 48 h). ELISA is proposed for use in the epidemiological investigation of the human tobacco risk.     

An evaluation of an enzyme immunoassay method for immunoreactive trypsin in dried blood spots

A novel monoclonal antibody based enzyme immunoassay (EIA) method for the measurement of the human cationic trypsinogen (NeoScreen, AGEN Biomedical Ltd., Acacia Ridge, Australia) in dried blood spots for the neonatal screening of cystic fibrosis was evaluated. The calibration standards provided as dried blood spots by AGEN are highly unstable and must be replaced with user prepared materials. Reference values from control individuals were obtained by parametric methods. A preliminary comparison with a polyclonal antibody based RIA method (Trypsik, SORIN Biomedica, Saluggia, Italy) was performed. Regression analysis between the RIA and the EIA methods gave a coefficient of correlation of 0.58 for RIA values less than 40 micrograms/L and of 0.77 for RIA values greater than or equal to 40 micrograms/L. Average CV of the within-run imprecision for the EIA method was 19.6% and for the RIA method 28.8%. CVs of the between-run imprecision at low, intermediate and high values for the EIA method were 23.7%, 15.8%, 15.6% and for the RIA method 20.6%, 14.4%, 11.2%. The diagnostic accuracy analyzed by a Receiver Operating Characteristics (ROC) curve of the RIA method gave a maximum accuracy of 190.9 while that of a simulated ROC curve for the EIA method was 193.0. We found that the precision and the diagnostic accuracy of the EIA method (AGEN) are equal to or better than those of one of the RIA methods.     

Evaluation of a monoclonal immunoenzymometric assay for alpha-fetoprotein

A monoclonal immunoenzymometric assay for alpha-fetoprotein was evaluated for detection and monitoring of hepatocellular carcinoma (HCC). We studied 1343 serum specimens from 759 patients with various neoplastic and non-neoplastic diseases. The interassay CV for this assay (M-AFP) ranged from 3.5 to 5.5%, with a minimum detectable concentration of 2.2 micrograms/L, as compared with 10 micrograms/L for a polyclonal (P-AFP) RIA for AFP. The calibration curve (0-300 micrograms/L) was linear, and serum dilutions paralleled it. The reference interval (0-9 micrograms/L) was established from data on 111 healthy subjects. Regression analysis of the AFP concentration (0-300 micrograms/L) of HCC patients obtained with the M-AFP assay (y) and the P-AFP RIA (x) yielded the equation y = (1.125)x - 0.52 (r = 0.9395, n = 165), with a considerable number of discrepant results for AFP less than 100 micrograms/L. By M-AFP immunoassay, AFP was above-normal (greater than 9 micrograms/L) in most HCC patients (80%), and to a lesser extent in other liver tumors (48%). AFP was within the normal reference interval for most patients with germ-cell tumors or benign liver disease and for other disease groups. For maximum diagnostic efficiency (90%) for HCC the decision level was increased to 100 micrograms of AFP per liter. Changes in serum AFP were correlated with changes in tumor volume in most HCC patients.     

脑脊液实验室检测在原发性中枢神经系统淋巴瘤诊断中的价值探讨

目的:探讨原发性中枢神经系统淋巴瘤（PCNSL）患者脑脊液的实验室检测指标在PCNSL诊断及鉴别诊断中的价值。方法:2017年12月至2020年1月在北京天坛医院诊断PCNSL的46例患者作为研究对象,选取同期就诊的中枢神经系统脱髓鞘疾病44例、中枢神经系统感染性疾病37例作为对照组,进行回顾性分析。比较3组患者的脑脊...     

Development and application of monoclonal antibodies to human cardiac myoglobin in a rapid fluorescence immunoassay.

Myoglobin (Mb) is considered a useful marker for early detection of myocardial infarction and for monitoring cardiac reperfusion after thrombolytic therapy. We developed eight monoclonal antibodies to human cardiac Mb, characterized their epitopic reactivity, and determined which combinations of the antibodies are useful in two-site immunoassays. We configured two of the monoclonal antibodies in a one-step, two-site particle concentration fluorescence immunoassay (PCFIA) for measurement of Mb. The PCFIA has rapid kinetics of reaction, being complete in 15 min, and has a linear analytical range of 20-675 micrograms/L for human Mb. Although the PCFIA has a high-dose "hook" effect, this is of no analytical importance at concentrations of Mb less than or equal to 148,000 micrograms/L. The assay is not subject to interference from icterus (bilirubin less than or equal to 360 mg/L), has no cross-reaction with hemoglobin (less than or equal to 42 g/L), and may be performed with either plasma or serum in approximately 1 h. The intra- and interassay imprecisions (CV) of the method are less than 10% for concentrations of Mb within the normal range and less than 4% at higher concentrations. A comparison of the PCFIA with a commercial radioimmunoassay showed that results of the two assays correlate well (PCFIA = 0.88 x RIA + 18, r = 0.990, n = 171).     

Urinary albumin excretion in normal subjects and in diabetic patients measured by a radioimmunoassay: methodological and clinical aspects

We have developed a radioimmunoassay method (RIA) to measure urinary albumin excretion. We determined the albumin excretion rate (AER) (micrograms/min) of 122 healthy subjects and 145 diabetic patients (115 type I, 30 type II). The results indicate that the RIA is sensitive (0.39 +/- 0.08 mg/L), precise (CV 5-8%), and gives reliable results on previously frozen urine samples. The distribution of the AER values in healthy subjects and diabetic patients was not normal. It was normalized by log or square-root transformation of the data. Seventy-three percent of diabetic patients lay within the normal range (0.6-10.6 micrograms/min). Twenty percent could be considered "at risk" to develop overt diabetic nephropathy because their albuminuria exceeded a threshold level of 15 micrograms/min chosen previously as the cutoff value for microalbuminuria. We found no correlation between AER and glycated hemoglobin, and only a weak correlation between AER and diabetes duration in type I diabetic patients.     

Immunofluorometric demonstration and quantification of placental protein 5 in the absence of pregnancy

This time-resolved immunofluorometric assay (IFMA) developed for measurement of placental protein 5 (PP5) involves two antibodies: a monoclonal anti-PP5 antibody attached to a solid phase and an europium(III) chelate-labeled polyclonal anti-PP5 antibody as a tracer. The measuring range is 0.05-100 micrograms/L and the detection limit is 20 times lower than that of a PP5 radioimmunoassay (RIA) performed with the same polyclonal antiserum. By IFMA, PP5 could be detected and quantified in all plasma and serum samples of nonpregnant and pregnant individuals, whereas PP5 was undetectable by RIA in serum of healthy men and nonpregnant women. The mean concentration of PP5 in sera from men was 0.43 micrograms/L (SD 0.13, range 0.19-0.75, n = 47) and in sera from nonpregnant women 0.49 micrograms/L (SD 0.19, range 0.20-0.90, n = 41). PP5 concentrations in serum showed no systematic variation during the menstrual cycle. In serum samples from 60 pregnant women the results obtained by IFMA and RIA correlated well (r = 0.97).     

Improved liquid-chromatographic determination of cyclosporine, with concomitant detection of a cell-bound metabolite

This unique extraction and isocratic "high-performance" liquid chromatographic method for measuring cyclosporine (CsA) in blood involves a Zorbax cyanopropyl analytical column maintained at 58 degrees C, with detection at 214 nm, and recycling of the water:acetonitrile mobile phase for improved long-term column stability and efficiency. Routinely, 1.0 mL of serum, plasma, or whole blood is diluted with water:acetonitrile (70:30) and applied to a disposable solid-phase cyanopropyl column to rapidly extract the drug and the internal standard cyclosporin D (CsD). Analytical recovery for this step averages 90% with whole blood and 98% with serum and plasma. Between-run CVs were 6.5 and 2.6% for means of 104 and 1128 micrograms/L, respectively. The standard curve is linear up to 1600 micrograms/L. The minimum detection limit is 10 to 15 micrograms/L. No interferences from endogenous substances or other drugs were found. In addition, a compound cross reacting with the Sandoz radioimmunoassay antibody was isolated from patients' samples with the present procedure and was tentatively identified as a CsA metabolite(s). It appears to be highly partitioned on blood cells, very little being detected in the serum or plasma. In a comparison with RIA, correlation coefficients were 0.828 and 0.652 for serum and whole blood, respectively. Results from a 12-h pharmacokinetic study in which different sample types were analyzed by RIA and liquid chromatography further exemplified major discrepancies between types of CsA determinations.     

Identification and analysis of nine metabolites of cyclosporine in whole blood by liquid chromatography. 2: Comparison of patients' results

We used a "high-performance" liquid-chromatographic assay [for parent cyclosporine (CsA) and nine metabolites] and a radioimmunoassay to detail the diversity of results among whole-blood samples from patients with transplanted organs. Heterogeneous populations of metabolites in samples collected just before the next dose of CsA were detected by HPLC, with CsA, M17, M1, or M8 predominating; M21, M203-218, MUNDF1, and M18 were detected in lesser amounts. Results by HPLC vs RIA for CsA or for individual metabolites vs CsA (or RIA) were diverse, with correlation coefficients (r) ranging from 0.058 to 0.933. RIA vs HPLC(sum of CsA + metabolites) gave the best comparison (slope = 0.931, y-intercept 14 micrograms/L, r = 0.933); but the scatter of data about the regression line remained significant (Sy/x = 132 micrograms/L). Most important, RIA/HPLC(CsA) vs HPLC(sum of metabolites) was remarkably poor (r = 0.222). A 12-h pharmacokinetic curve (for drug concentrations in a heart-transplant patient) displayed dissimilar times for peak concentrations of CsA and metabolites; each differed in the proportion (48% to 81% of peak concentration) eliminated from blood over the 12 h. These studies exemplify the utility of a more-inclusive, specific assay to monitor the diverse disposition of cyclosporines in patients and to demonstrate the errors associated with use of the RIA/HPLC ratio technique to predict metabolite concentrations.     

Autoreactive T lymphocytes in multiple sclerosis determined by antigen-induced secretion of interferon-gamma.

Multiple sclerosis (MS) is a disease with unknown cause characterized by inflammation and demyelination in the central nervous system. Although an autoimmune pathogenesis has been suggested, there are no conclusive data on the number of T cells autoreactive with myelin antigens in MS compared to controls. We showed that T lymphocytes secreting interferon-gamma in response to possible target autoantigens are severalfold more common among PBL mononuclear cells in patients with MS than in patients with aseptic meningitis and tension headache. On average T cells reactive with myelin basic protein (MBP), two different MBP peptides, or with proteolipid protein amounted to 2.7-5.2/10(5) PBL from MS patients. MBP-reactive T cells were still more frequent among mononuclear cells isolated from the cerebrospinal fluid (CSF; 185/10(5) CSF cells). We concluded that T cells reactive with myelin autoantigens are strongly increased in MS. This approach to detect them could allow definition of immunodominant T cell epitopes in individual MS patients, and thereby enable further development towards specific immunotherapy.     

Measurement of prostate-specific antigen by radioimmunoassay

We describe the performance of an RIA for the measurement of prostate-specific antigen (PA). Between-assay precision (CV) for control sera with various analyte concentrations was as follows: mean = 1.67 micrograms/L, CV = 7.1%; mean = 4.47 micrograms/L, CV = 5.6%; mean = 7.15 micrograms/L, CV = 5.5% (n = 19 each). Analytical recovery of PA (nine concentrations ranging from 2.3 to 21.1 micrograms/L) added to a serum pool averaged 101.8% (range 96.1 to 116.1%). Sensitivity (detection limit) of the RIA was 0.25 micrograms/L. Cross reactivity of prostatic acid phosphatase (PAP) in this assay was less than 0.022%. The mean percent B/B0 for 74 specimens from women was 98.9%, not statistically different from that for the zero standard. The normal reference interval for men was 0-2.7 micrograms/L ( 99th percentile), as established by assay of specimens from 276 apparently normal men. Measurement of PA and prostatic acid phosphatase in 205 consecutive serum specimens from patients with clinical evidence of prostate disease similarly placed patients into normal (98) or abnormal groups (54) in 152 cases. However, in 49 cases only the concentration of PA in serum was abnormal. Sequential measurement of both tissue markers in specimens from several patients who were undergoing therapy for prostate cancer appeared to provide supplemental information regarding treatment success.     

Purification of primary antibodies of the myelin basic protein antibody cascade from multiple sclerosis patients. Immunoreactivity studies with homologous and heterologous antigens.

Previous research has demonstrated a myelin basic protein (MBP) antibody cascade in cerebrospinal fluid (CSF) of multiple sclerosis (MS) patients. The purpose of this study was to determine whether primary antibodies to MBP (anti-MBP) reacted similarly with homologous (human) and heterologous (bovine and porcine) MBP. Myelin basic protein was prepared from central nervous system white matter of humans as well as bovine and porcine species. Immunoglobulin G (IgG) was purified by protein A-Sepharose affinity chromatography from concentrated CSF of MS patients with active or inactive disease or from non-MS controls. Antibodies to MBP were isolated from purified CSF IgG of MS patients with acute relapses by two-step antigen specific (MBP-Sepharose) affinity chromatography. Anti-MBP in the context of whole CSF, in purified CSF-IgG or as purified antibody, reacted identically with homologous and heterologous MBP. Kinetic studies of anti-MBP titers demonstrated that when anti-MBP was reacted in vitro with increasing amounts of homologous or heterologous MBP, the antibody was equally neutralized by either antigen. Neutralization of anti-MBP by homologous and heterologous MBP or their synthetic peptides may also be possible in vivo as a potential therapeutic tool.     

Concentrations and molecular forms of C-type natriuretic peptide in brain and cerebrospinal fluid.

To develop a radioimmunoassay (RIA) specific for human C-type natriuretic peptide (hCNP), we used a highly specific antiserum raised in rabbits. Quantitative inhibition tests with various natriuretic peptides demonstrated that the 50% inhibitory dose of hCNP was 15 fmol, whereas those of other natriuretic peptides were 10(5)-fold higher, indicating a specificity satisfactory for determining concentrations of hCNP in tissues. Using this antiserum, we detected immunoreactive hCNP (ir-hCNP) in various regions of human brain and spinal cord, as well as in cerebrospinal fluid (CSF). The ir-hCNP concentrations in human neural tissues were approximately 10-fold higher than those of immunoreactive human atrial natriuretic peptide (ir-hANP). The mean (+/- SD) concentration of ir-hCNP (72.0 +/- 17.8 ng/L) in CSF also was 10-fold higher than that of ir-hANP (5.2 +/- 2.1 ng/L). Using gel-permeation chromatography, we identified two molecular forms of ir-hCNP in brain and CSF: a 2-kDa form corresponding to mature hCNP, which is composed of 22 amino acid residues (hCNP-22), and a 5- to 6-kDa form corresponding to an N-terminally extended molecule (hCNP-53). The latter form was predominant in brain; the former was the main constituent of hCNP in CSF. These results support the hypothesis that hCNP is a major natriuretic peptide, is synthesized in human brain, and functions in human central nervous tissues.